66 PROCEEDINGS: BOSTON SOCIETY NATURAL HISTORY. 
nal syncytium is suggested, out of which the cortical nuclei are being 
organized into cells, while the mother branch cells are completely 
differentiated with the exception of their protoplasmic connection 
with the stem cell. 
I have never seen the grandmother stem cell divide, there being 
never more than one in a follicle (pi. 12, fig. 8b). The nucleus, how¬ 
ever, often presents appearances suggesting amitotic division (pi. 15, 
fig. 57a). Yet, aside from the granules resembling chromatin in the 
cytoplasm (pi. 15, fig. 56) I find nothing suggesting an origin of the 
cortical nuclei from the nuclei of the grandmother cell. 
The cortical nuclei are usually elongated, compressed, and irregular, 
and it may be possible that they multiply amitotically, their minute¬ 
ness preventing a definite determination of that matter. I have 
never seen any evidence of karyokinesis in these nuclei, though I 
have seen it in their descendants, the young cyst cells, soon to be con¬ 
sidered. On the other hand, the mother branch cells present beauti¬ 
ful mitotic figures (pi. 12, fig. 4; pi. 15, fig. 57a). It is necessary to 
consider very carefully whether the nucleus of the mother branch cell 
may not be merely developed cortical nuclei, and the whole mother 
branch cell in fact a developed cortical cell to be replaced by the latter 
as they develop. Facts do not sustain this hypothesis. The cortical 
nuclei not only surround the protoplasmic processes or branches, 
but extend also over and between the mother branch cells, where they 
are apt to be overlooked except in most favorable preparations. At 
the distal end of the mother branch cell, the cortical nuclei are gener¬ 
ally considerably larger (pi. 15, fig. 57a). 
Karyokinesis oj the mother branch cell .— During rest the mother 
branch cell is more or less conical, the apex being in all cases turned 
toward the grandmother stem cell, which is the point of attachment 
with the branch. Near the center of the proximal pointed pole there 
is often seen a body suggesting a centrosome and sphere with archo- 
plasm. 
The nucleus in the resting state is comparatively large, spherical, 
and rich in chromatin (pi. 15, fig. 57a, hr. c.). The latter is in the 
form of a network evenly distributed with spherical, deeply staining 
chromosomes near the periphery, and an irregular aggregation of 
deeply stained granules near the center — the only representative of a 
nucleolus. 
When division ensues, the nucleus is transformed into a compara- 
